We have increasing evidence that the protective effect of BCG varies widely between strains of M. tuberculosis [mostly of the Beijing family] isolated from patients in the USA, and those collected in the Western Cape of South Africa, the site of current TB vaccine trials. We have shown that BCG is only transiently protective against the US Beijing strains, but highly protective against two Western Cape strains in guinea pigs. The purpose of this R21 application therefore is to test [a] the efficacy of BCG against two larger panels representative of the two groups of isolates to see if our preliminary data is a collective phenomenon, and [b] to test the hypothesis that strains that have the capacity to resist the protective effects of BCG are able to do so because they potently induce regulatory T cells that inhibit this immunity, a hypothesis based on considerable new evidence from our laboratory. Not only do these studies have significant public health implications, notably for new vaccine trials, but if the Western Cape strains are as sensitive to BCG as our preliminary data suggests, they provide an explanation for the failure of the recent MVA85A trial. These studies will be conducted in our state of the art level-III facilities at CSU using conventional techniques at whic we are adept.